height_weight.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "height-weight.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/noahgift/regression-concepts/blob/master/height_weight.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "nuT_eom2-Ouv"
      },
      "source": [
        "# Height & Weight 25k Records"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "FcM872GDCLNo"
      },
      "source": [
        "* Original dataset from:  http://wiki.stat.ucla.edu/socr/index.php/SOCR_Data_Dinov_020108_HeightsWeights\n",
        "* 25,000 synthetic records of human heights and weights of 18 years old children"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "zy3h9Eo2Bqdu"
      },
      "source": [
        "## Ingest"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "BWWn4uKe-J-i"
      },
      "source": [
        "import pandas as pd"
      ],
      "execution_count": 5,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jGSvO0-G-bhT",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 208
        },
        "outputId": "341be29a-0dbe-4cd1-c701-8abeb83d5f21"
      },
      "source": [
        "df = pd.read_csv(\"https://raw.githubusercontent.com/noahgift/regression-concepts/master/height-weight-25k.csv\")\n",
        "df.head()"
      ],
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "\n",
              "  <div id=\"df-516f326f-4b2c-49ec-b146-fb155da36684\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Index</th>\n",
              "      <th>Height-Inches</th>\n",
              "      <th>Weight-Pounds</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>1</td>\n",
              "      <td>65.78331</td>\n",
              "      <td>112.9925</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>2</td>\n",
              "      <td>71.51521</td>\n",
              "      <td>136.4873</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>3</td>\n",
              "      <td>69.39874</td>\n",
              "      <td>153.0269</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>4</td>\n",
              "      <td>68.21660</td>\n",
              "      <td>142.3354</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>5</td>\n",
              "      <td>67.78781</td>\n",
              "      <td>144.2971</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-516f326f-4b2c-49ec-b146-fb155da36684')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-516f326f-4b2c-49ec-b146-fb155da36684 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-516f326f-4b2c-49ec-b146-fb155da36684');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ],
            "text/plain": [
              "   Index  Height-Inches  Weight-Pounds\n",
              "0      1       65.78331       112.9925\n",
              "1      2       71.51521       136.4873\n",
              "2      3       69.39874       153.0269\n",
              "3      4       68.21660       142.3354\n",
              "4      5       67.78781       144.2971"
            ]
          },
          "metadata": {},
          "execution_count": 6
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        ""
      ],
      "metadata": {
        "id": "cK9D2cuK06XL"
      },
      "execution_count": 6,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "GU5pPUgGBnoq"
      },
      "source": [
        "## EDA"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "26i11Jb6-hZh"
      },
      "source": [
        "### Scatterplot"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "fJT0DiOz-eYD"
      },
      "source": [
        "import seaborn as sns\n",
        "import numpy as np"
      ],
      "execution_count": 7,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "u1UBHvFz-k7T",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 442
        },
        "outputId": "8d8975c1-0a16-4f00-9aa1-11ed6a979adc"
      },
      "source": [
        "sns.lmplot(\"Height-Inches\", \"Weight-Pounds\", data=df)"
      ],
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/seaborn/_decorators.py:43: FutureWarning: Pass the following variables as keyword args: x, y. From version 0.12, the only valid positional argument will be `data`, and passing other arguments without an explicit keyword will result in an error or misinterpretation.\n",
            "  FutureWarning\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<seaborn.axisgrid.FacetGrid at 0x7fc4fec71fd0>"
            ]
          },
          "metadata": {},
          "execution_count": 8
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 360x360 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "wYp7hLRrB2w0"
      },
      "source": [
        "### Descriptive Statistics"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_bl3uvw-B69I",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 303
        },
        "outputId": "ad8761bd-ee81-48b6-8315-765c013e2f05"
      },
      "source": [
        "df.describe()"
      ],
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "\n",
              "  <div id=\"df-fd419fc6-4fc7-4467-a9da-7f60314a391f\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Index</th>\n",
              "      <th>Height-Inches</th>\n",
              "      <th>Weight-Pounds</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>count</th>\n",
              "      <td>25000.000000</td>\n",
              "      <td>25000.000000</td>\n",
              "      <td>25000.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>mean</th>\n",
              "      <td>12500.500000</td>\n",
              "      <td>67.993114</td>\n",
              "      <td>127.079421</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>std</th>\n",
              "      <td>7217.022701</td>\n",
              "      <td>1.901679</td>\n",
              "      <td>11.660898</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>min</th>\n",
              "      <td>1.000000</td>\n",
              "      <td>60.278360</td>\n",
              "      <td>78.014760</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>25%</th>\n",
              "      <td>6250.750000</td>\n",
              "      <td>66.704397</td>\n",
              "      <td>119.308675</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>50%</th>\n",
              "      <td>12500.500000</td>\n",
              "      <td>67.995700</td>\n",
              "      <td>127.157750</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>75%</th>\n",
              "      <td>18750.250000</td>\n",
              "      <td>69.272958</td>\n",
              "      <td>134.892850</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>max</th>\n",
              "      <td>25000.000000</td>\n",
              "      <td>75.152800</td>\n",
              "      <td>170.924000</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-fd419fc6-4fc7-4467-a9da-7f60314a391f')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-fd419fc6-4fc7-4467-a9da-7f60314a391f button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-fd419fc6-4fc7-4467-a9da-7f60314a391f');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ],
            "text/plain": [
              "              Index  Height-Inches  Weight-Pounds\n",
              "count  25000.000000   25000.000000   25000.000000\n",
              "mean   12500.500000      67.993114     127.079421\n",
              "std     7217.022701       1.901679      11.660898\n",
              "min        1.000000      60.278360      78.014760\n",
              "25%     6250.750000      66.704397     119.308675\n",
              "50%    12500.500000      67.995700     127.157750\n",
              "75%    18750.250000      69.272958     134.892850\n",
              "max    25000.000000      75.152800     170.924000"
            ]
          },
          "metadata": {},
          "execution_count": 9
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Allt4xOYMuDo"
      },
      "source": [
        "### Kernel Density Distribution"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ewqdEehUNzxs",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 497
        },
        "outputId": "fcacfb1d-c295-40be-ad73-3e2465d4bcd6"
      },
      "source": [
        "sns.jointplot(\"Height-Inches\", \"Weight-Pounds\", data=df, kind=\"kde\");"
      ],
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/seaborn/_decorators.py:43: FutureWarning: Pass the following variables as keyword args: x, y. From version 0.12, the only valid positional argument will be `data`, and passing other arguments without an explicit keyword will result in an error or misinterpretation.\n",
            "  FutureWarning\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x432 with 3 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "1VDfbgveBz_v"
      },
      "source": [
        "## Modeling "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "iQOlRnCSBH1Q"
      },
      "source": [
        "### Sklearn Regression Model"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "F-FQAkCb_Cg1"
      },
      "source": [
        "from sklearn.neighbors import KNeighborsRegressor\n",
        "from sklearn.model_selection import train_test_split"
      ],
      "execution_count": 11,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "I5iD7yijCxNH"
      },
      "source": [
        "#### Extract and Inspect Feature and Target"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "iGvWuX9fH7Gh"
      },
      "source": [
        "y = df['Weight-Pounds'].values #Target\n",
        "y = y.reshape(-1, 1)\n",
        "X = df['Height-Inches'].values #Feature(s)\n",
        "X = X.reshape(-1,1)"
      ],
      "execution_count": 12,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "aE1fd1CqIG8A",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "0d7c6b1a-1660-4d61-cdd2-60a9656e65ae"
      },
      "source": [
        "y.shape"
      ],
      "execution_count": 13,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(25000, 1)"
            ]
          },
          "metadata": {},
          "execution_count": 13
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "2b72FXk7IMwS"
      },
      "source": [
        "#### Split the Data"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LvgbbdeFIIDb",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "df48c7d1-fc9e-452a-f250-ff4b5c6fbb58"
      },
      "source": [
        "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2)\n",
        "print(X_train.shape, y_train.shape)\n",
        "print(X_test.shape, y_test.shape)"
      ],
      "execution_count": 14,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(20000, 1) (20000, 1)\n",
            "(5000, 1) (5000, 1)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "IE32wNfzIXtJ"
      },
      "source": [
        "#### Fit the model"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_WIPKSGMIZ4i"
      },
      "source": [
        "from sklearn.linear_model import LinearRegression\n",
        "lm = LinearRegression()\n",
        "model = lm.fit(X_train, y_train)\n",
        "y_predicted = lm.predict(X_test)"
      ],
      "execution_count": 32,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        ""
      ],
      "metadata": {
        "id": "rwcNG0jB1Qi0",
        "outputId": "d46bba0c-a581-4bb4-cb10-a0a47fc858fa",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 168
        }
      },
      "execution_count": 34,
      "outputs": [
        {
          "output_type": "error",
          "ename": "AttributeError",
          "evalue": "ignored",
          "traceback": [
            "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
            "\u001b[0;31mAttributeError\u001b[0m                            Traceback (most recent call last)",
            "\u001b[0;32m<ipython-input-34-36d0b8008a37>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0my_predicted\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m__dict__\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
            "\u001b[0;31mAttributeError\u001b[0m: 'numpy.ndarray' object has no attribute '__dict__'"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "REx74VwVIdDV"
      },
      "source": [
        "#### Print accuracy of Linear Regression Model"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Mo1Gl9T5I83s",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "90c0b92c-4fa9-4456-f556-c70ae6b128ba"
      },
      "source": [
        "from sklearn.metrics import mean_squared_error\n",
        "from math import sqrt\n",
        "\n",
        "#RMSE Root Mean Squared Error\n",
        "rms = sqrt(mean_squared_error(y_predicted, y_test))\n",
        "rms"
      ],
      "execution_count": 16,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "9.966074183159119"
            ]
          },
          "metadata": {},
          "execution_count": 16
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "TZDGFJXRJGDp"
      },
      "source": [
        "#### Plot Predicted Height versus Actual Height"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "eq0E7pBCJI2p",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        },
        "outputId": "699718c4-4722-4301-ab17-4fb87ec0f49c"
      },
      "source": [
        "import matplotlib.pyplot as plt\n",
        "_, ax = plt.subplots()\n",
        "\n",
        "ax.scatter(x = range(0, y_test.size), y=y_test, c = 'blue', label = 'Actual', alpha = 0.5)\n",
        "ax.scatter(x = range(0, y_predicted.size), y=y_predicted, c = 'red', label = 'Predicted', alpha = 0.5)\n",
        "\n",
        "plt.title('Actual Height vs Predicted Height')\n",
        "plt.xlabel('Weight')\n",
        "plt.ylabel('Height')\n",
        "plt.legend()\n",
        "plt.show()"
      ],
      "execution_count": 17,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "vZCFefGARL0d"
      },
      "source": [
        "### Regression with PyTorch"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "9rEbjaWpTH8w"
      },
      "source": [
        "convert data to float"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "L6Tx0tobTKLf"
      },
      "source": [
        "# Training Data\n",
        "x_train = np.array(X_train, dtype=np.float32)\n",
        "x_train = x_train.reshape(-1, 1)\n",
        "y_train = np.array(y_train, dtype=np.float32)\n",
        "y_train = y_train.reshape(-1, 1)\n",
        "\n",
        "# Test Data\n",
        "x_test = np.array(X_test, dtype=np.float32)\n",
        "x_test = x_test.reshape(-1, 1)\n",
        "y_test = np.array(y_test, dtype=np.float32)\n",
        "y_test = y_test.reshape(-1, 1)"
      ],
      "execution_count": 18,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "7bHW7JZCRt7b"
      },
      "source": [
        "import torch\n",
        "from torch.autograd import Variable\n",
        "\n",
        "class linearRegression(torch.nn.Module):\n",
        "    def __init__(self, inputSize, outputSize):\n",
        "        super(linearRegression, self).__init__()\n",
        "        self.linear = torch.nn.Linear(inputSize, outputSize)\n",
        "\n",
        "    def forward(self, x):\n",
        "        out = self.linear(x)\n",
        "        return out"
      ],
      "execution_count": 19,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "wNDwck1GR0bd"
      },
      "source": [
        "Create a model with CUDA enabled"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1r0tQp7NR2Ou",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "b19338d8-6627-4bc4-e9d8-889c8ede56e5"
      },
      "source": [
        "inputDim = 1        # takes variable 'x' \n",
        "outputDim = 1       # takes variable 'y'\n",
        "learningRate = 0.0001 \n",
        "epochs = 1000\n",
        "\n",
        "model = linearRegression(inputDim, outputDim)\n",
        "model.cuda()"
      ],
      "execution_count": 20,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "linearRegression(\n",
              "  (linear): Linear(in_features=1, out_features=1, bias=True)\n",
              ")"
            ]
          },
          "metadata": {},
          "execution_count": 20
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "mUJNmPY-R7Vv"
      },
      "source": [
        "Create Stochastic Gradient Descent and Loss Function"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4VkOQhhtR96R"
      },
      "source": [
        "criterion = torch.nn.MSELoss() \n",
        "optimizer = torch.optim.SGD(model.parameters(), lr=learningRate)"
      ],
      "execution_count": 21,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "pIWXOa8-SNRL"
      },
      "source": [
        "train the model"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "BnO01W38SO6l",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "b98cce29-14a0-484f-a8c7-0f0ab2bd5b08"
      },
      "source": [
        "for epoch in range(epochs):\n",
        "    inputs = Variable(torch.from_numpy(x_train).cuda())\n",
        "    labels = Variable(torch.from_numpy(y_train).cuda())\n",
        "    optimizer.zero_grad()\n",
        "    outputs = model(inputs)\n",
        "    loss = criterion(outputs, labels)\n",
        "    print(loss)\n",
        "    # get gradients w.r.t to parameters\n",
        "    loss.backward()\n",
        "    # update parameters\n",
        "    optimizer.step()\n",
        "    print('epoch {}, loss {}'.format(epoch, loss.item()))"
      ],
      "execution_count": 22,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "tensor(20903.9648, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 0, loss 20903.96484375\n",
            "tensor(224.0141, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 1, loss 224.01409912109375\n",
            "tensor(108.2931, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 2, loss 108.29308319091797\n",
            "tensor(107.6455, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 3, loss 107.64552307128906\n",
            "tensor(107.6419, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 4, loss 107.64189910888672\n",
            "tensor(107.6419, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 5, loss 107.6418685913086\n",
            "tensor(107.6419, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 6, loss 107.64188385009766\n",
            "tensor(107.6419, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 7, loss 107.6418685913086\n",
            "tensor(107.6419, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 8, loss 107.64188385009766\n",
            "tensor(107.6419, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 9, loss 107.64188385009766\n",
            "tensor(107.6419, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 10, loss 107.6418685913086\n",
            "tensor(107.6419, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 11, loss 107.6418685913086\n",
            "tensor(107.6419, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 12, loss 107.6418685913086\n",
            "tensor(107.6419, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 13, loss 107.6418685913086\n",
            "tensor(107.6419, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 14, loss 107.6418685913086\n",
            "tensor(107.6419, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 15, loss 107.64186096191406\n",
            "tensor(107.6419, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 16, loss 107.64186096191406\n",
            "tensor(107.6419, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 17, loss 107.64186096191406\n",
            "tensor(107.6419, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 18, loss 107.64186096191406\n",
            "tensor(107.6419, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 19, loss 107.64186096191406\n",
            "tensor(107.6419, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 20, loss 107.64186096191406\n",
            "tensor(107.6419, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 21, loss 107.64186096191406\n",
            "tensor(107.6419, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 22, loss 107.64186096191406\n",
            "tensor(107.6418, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 23, loss 107.641845703125\n",
            "tensor(107.6419, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 24, loss 107.64186096191406\n",
            "tensor(107.6418, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 25, loss 107.641845703125\n",
            "tensor(107.6418, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 26, loss 107.641845703125\n",
            "tensor(107.6418, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 27, loss 107.641845703125\n",
            "tensor(107.6418, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 28, loss 107.641845703125\n",
            "tensor(107.6418, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 29, loss 107.64183807373047\n",
            "tensor(107.6418, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 30, loss 107.64183807373047\n",
            "tensor(107.6418, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 31, loss 107.64183807373047\n",
            "tensor(107.6418, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 32, loss 107.64183807373047\n",
            "tensor(107.6418, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 33, loss 107.64183807373047\n",
            "tensor(107.6418, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 34, loss 107.64183807373047\n",
            "tensor(107.6418, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 35, loss 107.64183807373047\n",
            "tensor(107.6418, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 36, loss 107.64183807373047\n",
            "tensor(107.6418, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 37, loss 107.64183807373047\n",
            "tensor(107.6418, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 38, loss 107.6418228149414\n",
            "tensor(107.6418, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 39, loss 107.6418228149414\n",
            "tensor(107.6418, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 40, loss 107.6418228149414\n",
            "tensor(107.6418, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 41, loss 107.6418228149414\n",
            "tensor(107.6418, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 42, loss 107.6418228149414\n",
            "tensor(107.6418, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 43, loss 107.64180755615234\n",
            "tensor(107.6418, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 44, loss 107.64180755615234\n",
            "tensor(107.6418, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 45, loss 107.64180755615234\n",
            "tensor(107.6418, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 46, loss 107.64180755615234\n",
            "tensor(107.6418, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 47, loss 107.64180755615234\n",
            "tensor(107.6418, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 48, loss 107.64180755615234\n",
            "tensor(107.6418, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 49, loss 107.64180755615234\n",
            "tensor(107.6418, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 50, loss 107.64180755615234\n",
            "tensor(107.6418, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 51, loss 107.64180755615234\n",
            "tensor(107.6418, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 52, loss 107.64180755615234\n",
            "tensor(107.6418, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 53, loss 107.64179992675781\n",
            "tensor(107.6418, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 54, loss 107.64179992675781\n",
            "tensor(107.6418, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 55, loss 107.64179992675781\n",
            "tensor(107.6418, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 56, loss 107.64179992675781\n",
            "tensor(107.6418, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 57, loss 107.64178466796875\n",
            "tensor(107.6418, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 58, loss 107.64178466796875\n",
            "tensor(107.6418, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 59, loss 107.64178466796875\n",
            "tensor(107.6418, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 60, loss 107.64178466796875\n",
            "tensor(107.6418, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 61, loss 107.64178466796875\n",
            "tensor(107.6418, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 62, loss 107.64178466796875\n",
            "tensor(107.6418, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 63, loss 107.64178466796875\n",
            "tensor(107.6418, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 64, loss 107.64178466796875\n",
            "tensor(107.6418, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 65, loss 107.64178466796875\n",
            "tensor(107.6418, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 66, loss 107.64178466796875\n",
            "tensor(107.6418, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 67, loss 107.64178466796875\n",
            "tensor(107.6418, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 68, loss 107.64178466796875\n",
            "tensor(107.6418, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 69, loss 107.64176940917969\n",
            "tensor(107.6418, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 70, loss 107.64176940917969\n",
            "tensor(107.6418, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 71, loss 107.64176177978516\n",
            "tensor(107.6418, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 72, loss 107.64176177978516\n",
            "tensor(107.6418, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 73, loss 107.64176177978516\n",
            "tensor(107.6418, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 74, loss 107.64176177978516\n",
            "tensor(107.6418, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 75, loss 107.64176177978516\n",
            "tensor(107.6418, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 76, loss 107.64176177978516\n",
            "tensor(107.6418, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 77, loss 107.64176177978516\n",
            "tensor(107.6418, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 78, loss 107.64176177978516\n",
            "tensor(107.6418, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 79, loss 107.64176177978516\n",
            "tensor(107.6418, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 80, loss 107.64176177978516\n",
            "tensor(107.6418, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 81, loss 107.64176177978516\n",
            "tensor(107.6418, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 82, loss 107.64176177978516\n",
            "tensor(107.6418, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 83, loss 107.64176177978516\n",
            "tensor(107.6417, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 84, loss 107.6417465209961\n",
            "tensor(107.6417, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 85, loss 107.6417465209961\n",
            "tensor(107.6417, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 86, loss 107.6417465209961\n",
            "tensor(107.6417, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 87, loss 107.64173126220703\n",
            "tensor(107.6417, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 88, loss 107.6417465209961\n",
            "tensor(107.6417, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 89, loss 107.64173126220703\n",
            "tensor(107.6417, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 90, loss 107.64173126220703\n",
            "tensor(107.6417, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 91, loss 107.64173126220703\n",
            "tensor(107.6417, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 92, loss 107.64173126220703\n",
            "tensor(107.6417, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 93, loss 107.64173126220703\n",
            "tensor(107.6417, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 94, loss 107.64173126220703\n",
            "tensor(107.6417, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 95, loss 107.64173126220703\n",
            "tensor(107.6417, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 96, loss 107.64173126220703\n",
            "tensor(107.6417, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 97, loss 107.64173126220703\n",
            "tensor(107.6417, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 98, loss 107.6417236328125\n",
            "tensor(107.6417, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 99, loss 107.6417236328125\n",
            "tensor(107.6417, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 100, loss 107.6417236328125\n",
            "tensor(107.6417, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 101, loss 107.6417236328125\n",
            "tensor(107.6417, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 102, loss 107.64170837402344\n",
            "tensor(107.6417, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 103, loss 107.64170837402344\n",
            "tensor(107.6417, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 104, loss 107.6417236328125\n",
            "tensor(107.6417, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 105, loss 107.64170837402344\n",
            "tensor(107.6417, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 106, loss 107.64170837402344\n",
            "tensor(107.6417, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 107, loss 107.64170837402344\n",
            "tensor(107.6417, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 108, loss 107.64170837402344\n",
            "tensor(107.6417, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 109, loss 107.64170837402344\n",
            "tensor(107.6417, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 110, loss 107.64170837402344\n",
            "tensor(107.6417, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 111, loss 107.64170837402344\n",
            "tensor(107.6417, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 112, loss 107.6417007446289\n",
            "tensor(107.6417, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 113, loss 107.64170837402344\n",
            "tensor(107.6417, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 114, loss 107.6417007446289\n",
            "tensor(107.6417, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 115, loss 107.6417007446289\n",
            "tensor(107.6417, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 116, loss 107.6417007446289\n",
            "tensor(107.6417, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 117, loss 107.64168548583984\n",
            "tensor(107.6417, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 118, loss 107.6417007446289\n",
            "tensor(107.6417, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 119, loss 107.64168548583984\n",
            "tensor(107.6417, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 120, loss 107.64168548583984\n",
            "tensor(107.6417, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 121, loss 107.64168548583984\n",
            "tensor(107.6417, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 122, loss 107.64168548583984\n",
            "tensor(107.6417, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 123, loss 107.64168548583984\n",
            "tensor(107.6417, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 124, loss 107.64168548583984\n",
            "tensor(107.6417, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 125, loss 107.64168548583984\n",
            "tensor(107.6417, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 126, loss 107.64168548583984\n",
            "tensor(107.6417, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 127, loss 107.64167022705078\n",
            "tensor(107.6417, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 128, loss 107.64167022705078\n",
            "tensor(107.6417, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 129, loss 107.64167022705078\n",
            "tensor(107.6417, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 130, loss 107.64167022705078\n",
            "tensor(107.6417, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 131, loss 107.64166259765625\n",
            "tensor(107.6417, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 132, loss 107.64167022705078\n",
            "tensor(107.6417, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 133, loss 107.64167022705078\n",
            "tensor(107.6417, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 134, loss 107.64166259765625\n",
            "tensor(107.6417, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 135, loss 107.64166259765625\n",
            "tensor(107.6417, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 136, loss 107.64166259765625\n",
            "tensor(107.6417, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 137, loss 107.64166259765625\n",
            "tensor(107.6417, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 138, loss 107.64166259765625\n",
            "tensor(107.6417, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 139, loss 107.64166259765625\n",
            "tensor(107.6417, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 140, loss 107.64166259765625\n",
            "tensor(107.6416, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 141, loss 107.64164733886719\n",
            "tensor(107.6416, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 142, loss 107.64164733886719\n",
            "tensor(107.6416, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 143, loss 107.64164733886719\n",
            "tensor(107.6416, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 144, loss 107.64164733886719\n",
            "tensor(107.6416, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 145, loss 107.64164733886719\n",
            "tensor(107.6416, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 146, loss 107.64163208007812\n",
            "tensor(107.6416, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 147, loss 107.64164733886719\n",
            "tensor(107.6416, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 148, loss 107.64163208007812\n",
            "tensor(107.6416, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 149, loss 107.64163208007812\n",
            "tensor(107.6416, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 150, loss 107.64163208007812\n",
            "tensor(107.6416, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 151, loss 107.64163208007812\n",
            "tensor(107.6416, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 152, loss 107.64163208007812\n",
            "tensor(107.6416, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 153, loss 107.64163208007812\n",
            "tensor(107.6416, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 154, loss 107.6416244506836\n",
            "tensor(107.6416, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 155, loss 107.6416244506836\n",
            "tensor(107.6416, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 156, loss 107.6416244506836\n",
            "tensor(107.6416, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 157, loss 107.6416244506836\n",
            "tensor(107.6416, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 158, loss 107.6416244506836\n",
            "tensor(107.6416, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 159, loss 107.6416244506836\n",
            "tensor(107.6416, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 160, loss 107.6416244506836\n",
            "tensor(107.6416, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 161, loss 107.6416244506836\n",
            "tensor(107.6416, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 162, loss 107.64160919189453\n",
            "tensor(107.6416, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 163, loss 107.64160919189453\n",
            "tensor(107.6416, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 164, loss 107.64160919189453\n",
            "tensor(107.6416, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 165, loss 107.64160919189453\n",
            "tensor(107.6416, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 166, loss 107.64160919189453\n",
            "tensor(107.6416, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 167, loss 107.64160919189453\n",
            "tensor(107.6416, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 168, loss 107.64160919189453\n",
            "tensor(107.6416, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 169, loss 107.64160919189453\n",
            "tensor(107.6416, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 170, loss 107.64160919189453\n",
            "tensor(107.6416, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 171, loss 107.64160919189453\n",
            "tensor(107.6416, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 172, loss 107.64159393310547\n",
            "tensor(107.6416, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 173, loss 107.64159393310547\n",
            "tensor(107.6416, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 174, loss 107.64159393310547\n",
            "tensor(107.6416, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 175, loss 107.64159393310547\n",
            "tensor(107.6416, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 176, loss 107.64158630371094\n",
            "tensor(107.6416, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 177, loss 107.64158630371094\n",
            "tensor(107.6416, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 178, loss 107.64158630371094\n",
            "tensor(107.6416, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 179, loss 107.64158630371094\n",
            "tensor(107.6416, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 180, loss 107.64158630371094\n",
            "tensor(107.6416, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 181, loss 107.64158630371094\n",
            "tensor(107.6416, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 182, loss 107.64158630371094\n",
            "tensor(107.6416, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 183, loss 107.64158630371094\n",
            "tensor(107.6416, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 184, loss 107.64158630371094\n",
            "tensor(107.6416, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 185, loss 107.64157104492188\n",
            "tensor(107.6416, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 186, loss 107.64157104492188\n",
            "tensor(107.6416, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 187, loss 107.64157104492188\n",
            "tensor(107.6416, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 188, loss 107.64157104492188\n",
            "tensor(107.6416, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 189, loss 107.64157104492188\n",
            "tensor(107.6416, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 190, loss 107.64157104492188\n",
            "tensor(107.6416, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 191, loss 107.64156341552734\n",
            "tensor(107.6416, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 192, loss 107.64156341552734\n",
            "tensor(107.6416, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 193, loss 107.64156341552734\n",
            "tensor(107.6416, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 194, loss 107.64156341552734\n",
            "tensor(107.6416, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 195, loss 107.64156341552734\n",
            "tensor(107.6416, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 196, loss 107.64156341552734\n",
            "tensor(107.6416, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 197, loss 107.64156341552734\n",
            "tensor(107.6415, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 198, loss 107.64154815673828\n",
            "tensor(107.6416, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 199, loss 107.64156341552734\n",
            "tensor(107.6415, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 200, loss 107.64154815673828\n",
            "tensor(107.6415, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 201, loss 107.64154815673828\n",
            "tensor(107.6415, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 202, loss 107.64154815673828\n",
            "tensor(107.6415, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 203, loss 107.64154815673828\n",
            "tensor(107.6415, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 204, loss 107.64154815673828\n",
            "tensor(107.6415, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 205, loss 107.64153289794922\n",
            "tensor(107.6415, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 206, loss 107.64153289794922\n",
            "tensor(107.6415, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 207, loss 107.64153289794922\n",
            "tensor(107.6415, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 208, loss 107.64153289794922\n",
            "tensor(107.6415, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 209, loss 107.64153289794922\n",
            "tensor(107.6415, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 210, loss 107.64153289794922\n",
            "tensor(107.6415, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 211, loss 107.64153289794922\n",
            "tensor(107.6415, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 212, loss 107.64153289794922\n",
            "tensor(107.6415, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 213, loss 107.64153289794922\n",
            "tensor(107.6415, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 214, loss 107.64152526855469\n",
            "tensor(107.6415, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 215, loss 107.64152526855469\n",
            "tensor(107.6415, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 216, loss 107.64152526855469\n",
            "tensor(107.6415, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 217, loss 107.64152526855469\n",
            "tensor(107.6415, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 218, loss 107.64152526855469\n",
            "tensor(107.6415, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 219, loss 107.64152526855469\n",
            "tensor(107.6415, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 220, loss 107.64152526855469\n",
            "tensor(107.6415, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 221, loss 107.64152526855469\n",
            "tensor(107.6415, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 222, loss 107.64151000976562\n",
            "tensor(107.6415, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 223, loss 107.64151000976562\n",
            "tensor(107.6415, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 224, loss 107.64151000976562\n",
            "tensor(107.6415, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 225, loss 107.64151000976562\n",
            "tensor(107.6415, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 226, loss 107.64151000976562\n",
            "tensor(107.6415, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 227, loss 107.64151000976562\n",
            "tensor(107.6415, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 228, loss 107.64149475097656\n",
            "tensor(107.6415, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 229, loss 107.64151000976562\n",
            "tensor(107.6415, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 230, loss 107.64149475097656\n",
            "tensor(107.6415, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 231, loss 107.64149475097656\n",
            "tensor(107.6415, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 232, loss 107.64149475097656\n",
            "tensor(107.6415, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 233, loss 107.64149475097656\n",
            "tensor(107.6415, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 234, loss 107.64149475097656\n",
            "tensor(107.6415, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 235, loss 107.64149475097656\n",
            "tensor(107.6415, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 236, loss 107.64148712158203\n",
            "tensor(107.6415, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 237, loss 107.64148712158203\n",
            "tensor(107.6415, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 238, loss 107.64148712158203\n",
            "tensor(107.6415, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 239, loss 107.64148712158203\n",
            "tensor(107.6415, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 240, loss 107.64148712158203\n",
            "tensor(107.6415, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 241, loss 107.64148712158203\n",
            "tensor(107.6415, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 242, loss 107.64148712158203\n",
            "tensor(107.6415, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 243, loss 107.64148712158203\n",
            "tensor(107.6415, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 244, loss 107.64148712158203\n",
            "tensor(107.6415, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 245, loss 107.64147186279297\n",
            "tensor(107.6415, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 246, loss 107.64147186279297\n",
            "tensor(107.6415, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 247, loss 107.64147186279297\n",
            "tensor(107.6415, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 248, loss 107.64147186279297\n",
            "tensor(107.6415, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 249, loss 107.6414566040039\n",
            "tensor(107.6415, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 250, loss 107.6414566040039\n",
            "tensor(107.6415, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 251, loss 107.6414566040039\n",
            "tensor(107.6415, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 252, loss 107.64147186279297\n",
            "tensor(107.6415, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 253, loss 107.6414566040039\n",
            "tensor(107.6415, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 254, loss 107.6414566040039\n",
            "tensor(107.6415, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 255, loss 107.6414566040039\n",
            "tensor(107.6415, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 256, loss 107.6414566040039\n",
            "tensor(107.6415, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 257, loss 107.6414566040039\n",
            "tensor(107.6414, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 258, loss 107.64144897460938\n",
            "tensor(107.6414, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 259, loss 107.64144897460938\n",
            "tensor(107.6414, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 260, loss 107.64144897460938\n",
            "tensor(107.6414, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 261, loss 107.64144897460938\n",
            "tensor(107.6414, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 262, loss 107.64144897460938\n",
            "tensor(107.6414, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 263, loss 107.64143371582031\n",
            "tensor(107.6414, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 264, loss 107.64144897460938\n",
            "tensor(107.6414, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 265, loss 107.64143371582031\n",
            "tensor(107.6414, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 266, loss 107.64143371582031\n",
            "tensor(107.6414, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 267, loss 107.64143371582031\n",
            "tensor(107.6414, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 268, loss 107.64143371582031\n",
            "tensor(107.6414, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 269, loss 107.64143371582031\n",
            "tensor(107.6414, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 270, loss 107.64143371582031\n",
            "tensor(107.6414, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 271, loss 107.64143371582031\n",
            "tensor(107.6414, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 272, loss 107.64143371582031\n",
            "tensor(107.6414, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 273, loss 107.64142608642578\n",
            "tensor(107.6414, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 274, loss 107.64142608642578\n",
            "tensor(107.6414, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 275, loss 107.64142608642578\n",
            "tensor(107.6414, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 276, loss 107.64142608642578\n",
            "tensor(107.6414, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 277, loss 107.64141082763672\n",
            "tensor(107.6414, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 278, loss 107.64142608642578\n",
            "tensor(107.6414, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 279, loss 107.64141082763672\n",
            "tensor(107.6414, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 280, loss 107.64141082763672\n",
            "tensor(107.6414, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 281, loss 107.64141082763672\n",
            "tensor(107.6414, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 282, loss 107.64141082763672\n",
            "tensor(107.6414, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 283, loss 107.64141082763672\n",
            "tensor(107.6414, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 284, loss 107.64141082763672\n",
            "tensor(107.6414, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 285, loss 107.64139556884766\n",
            "tensor(107.6414, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 286, loss 107.64139556884766\n",
            "tensor(107.6414, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 287, loss 107.64139556884766\n",
            "tensor(107.6414, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 288, loss 107.64139556884766\n",
            "tensor(107.6414, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 289, loss 107.64139556884766\n",
            "tensor(107.6414, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 290, loss 107.64139556884766\n",
            "tensor(107.6414, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 291, loss 107.64139556884766\n",
            "tensor(107.6414, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 292, loss 107.64139556884766\n",
            "tensor(107.6414, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 293, loss 107.64139556884766\n",
            "tensor(107.6414, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 294, loss 107.64139556884766\n",
            "tensor(107.6414, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 295, loss 107.64139556884766\n",
            "tensor(107.6414, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 296, loss 107.64138793945312\n",
            "tensor(107.6414, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 297, loss 107.64139556884766\n",
            "tensor(107.6414, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 298, loss 107.64138793945312\n",
            "tensor(107.6414, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 299, loss 107.64137268066406\n",
            "tensor(107.6414, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 300, loss 107.64138793945312\n",
            "tensor(107.6414, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 301, loss 107.64137268066406\n",
            "tensor(107.6414, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 302, loss 107.64137268066406\n",
            "tensor(107.6414, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 303, loss 107.64137268066406\n",
            "tensor(107.6414, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 304, loss 107.64137268066406\n",
            "tensor(107.6414, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 305, loss 107.64137268066406\n",
            "tensor(107.6414, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 306, loss 107.64137268066406\n",
            "tensor(107.6414, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 307, loss 107.64137268066406\n",
            "tensor(107.6414, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 308, loss 107.64137268066406\n",
            "tensor(107.6414, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 309, loss 107.64137268066406\n",
            "tensor(107.6414, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 310, loss 107.64137268066406\n",
            "tensor(107.6414, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 311, loss 107.641357421875\n",
            "tensor(107.6414, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 312, loss 107.641357421875\n",
            "tensor(107.6414, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 313, loss 107.641357421875\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 314, loss 107.64134979248047\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 315, loss 107.64134979248047\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 316, loss 107.64134979248047\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 317, loss 107.64134979248047\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 318, loss 107.64134979248047\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 319, loss 107.64134979248047\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 320, loss 107.64134979248047\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 321, loss 107.64134979248047\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 322, loss 107.64134979248047\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 323, loss 107.64134979248047\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 324, loss 107.64134979248047\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 325, loss 107.64134979248047\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 326, loss 107.64134979248047\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 327, loss 107.64134979248047\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 328, loss 107.6413345336914\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 329, loss 107.6413345336914\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 330, loss 107.64131927490234\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 331, loss 107.64131927490234\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 332, loss 107.64131927490234\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 333, loss 107.64131927490234\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 334, loss 107.64131927490234\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 335, loss 107.64131927490234\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 336, loss 107.64131927490234\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 337, loss 107.64131927490234\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 338, loss 107.64131927490234\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 339, loss 107.64131927490234\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 340, loss 107.64131927490234\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 341, loss 107.64131927490234\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 342, loss 107.64131164550781\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 343, loss 107.64129638671875\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 344, loss 107.64129638671875\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 345, loss 107.64129638671875\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 346, loss 107.64129638671875\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 347, loss 107.64129638671875\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 348, loss 107.64129638671875\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 349, loss 107.64129638671875\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 350, loss 107.64129638671875\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 351, loss 107.64129638671875\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 352, loss 107.64129638671875\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 353, loss 107.64129638671875\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 354, loss 107.64129638671875\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 355, loss 107.64129638671875\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 356, loss 107.64128112792969\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 357, loss 107.64127349853516\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 358, loss 107.64128112792969\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 359, loss 107.64127349853516\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 360, loss 107.64127349853516\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 361, loss 107.64127349853516\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 362, loss 107.64127349853516\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 363, loss 107.64127349853516\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 364, loss 107.64127349853516\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 365, loss 107.64127349853516\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 366, loss 107.64127349853516\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 367, loss 107.64127349853516\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 368, loss 107.64127349853516\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 369, loss 107.6412582397461\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 370, loss 107.64127349853516\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 371, loss 107.6412582397461\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 372, loss 107.64125061035156\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 373, loss 107.64125061035156\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 374, loss 107.64125061035156\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 375, loss 107.64125061035156\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 376, loss 107.64125061035156\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 377, loss 107.64125061035156\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 378, loss 107.64125061035156\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 379, loss 107.64125061035156\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 380, loss 107.64125061035156\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 381, loss 107.64125061035156\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 382, loss 107.64125061035156\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 383, loss 107.64125061035156\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 384, loss 107.64125061035156\n",
            "tensor(107.6412, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 385, loss 107.6412353515625\n",
            "tensor(107.6413, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 386, loss 107.64125061035156\n",
            "tensor(107.6412, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 387, loss 107.64122009277344\n",
            "tensor(107.6412, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 388, loss 107.64122009277344\n",
            "tensor(107.6412, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 389, loss 107.64122009277344\n",
            "tensor(107.6412, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 390, loss 107.64122009277344\n",
            "tensor(107.6412, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 391, loss 107.64122009277344\n",
            "tensor(107.6412, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 392, loss 107.64122009277344\n",
            "tensor(107.6412, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 393, loss 107.64122009277344\n",
            "tensor(107.6412, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 394, loss 107.64122009277344\n",
            "tensor(107.6412, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 395, loss 107.64122009277344\n",
            "tensor(107.6412, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 396, loss 107.64122009277344\n",
            "tensor(107.6412, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 397, loss 107.64122009277344\n",
            "tensor(107.6412, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 398, loss 107.64122009277344\n",
            "tensor(107.6412, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 399, loss 107.64122009277344\n",
            "tensor(107.6412, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 400, loss 107.64122009277344\n",
            "tensor(107.6412, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 401, loss 107.6412124633789\n",
            "tensor(107.6412, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 402, loss 107.64119720458984\n",
            "tensor(107.6412, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 403, loss 107.64119720458984\n",
            "tensor(107.6412, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 404, loss 107.64119720458984\n",
            "tensor(107.6412, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 405, loss 107.64119720458984\n",
            "tensor(107.6412, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 406, loss 107.64119720458984\n",
            "tensor(107.6412, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 407, loss 107.64119720458984\n",
            "tensor(107.6412, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 408, loss 107.64119720458984\n",
            "tensor(107.6412, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 409, loss 107.64119720458984\n",
            "tensor(107.6412, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 410, loss 107.64119720458984\n",
            "tensor(107.6412, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 411, loss 107.64119720458984\n",
            "tensor(107.6412, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 412, loss 107.64119720458984\n",
            "tensor(107.6412, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 413, loss 107.64119720458984\n",
            "tensor(107.6412, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 414, loss 107.64119720458984\n",
            "tensor(107.6412, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 415, loss 107.64118194580078\n",
            "tensor(107.6412, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 416, loss 107.64119720458984\n",
            "tensor(107.6412, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 417, loss 107.64118194580078\n",
            "tensor(107.6412, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 418, loss 107.64117431640625\n",
            "tensor(107.6412, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 419, loss 107.64117431640625\n",
            "tensor(107.6412, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 420, loss 107.64117431640625\n",
            "tensor(107.6412, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 421, loss 107.64117431640625\n",
            "tensor(107.6412, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 422, loss 107.64117431640625\n",
            "tensor(107.6412, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 423, loss 107.64117431640625\n",
            "tensor(107.6412, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 424, loss 107.64117431640625\n",
            "tensor(107.6412, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 425, loss 107.64117431640625\n",
            "tensor(107.6412, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 426, loss 107.64117431640625\n",
            "tensor(107.6412, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 427, loss 107.64117431640625\n",
            "tensor(107.6412, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 428, loss 107.64117431640625\n",
            "tensor(107.6412, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 429, loss 107.64117431640625\n",
            "tensor(107.6412, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 430, loss 107.64115905761719\n",
            "tensor(107.6412, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 431, loss 107.64115905761719\n",
            "tensor(107.6411, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 432, loss 107.64114379882812\n",
            "tensor(107.6411, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 433, loss 107.64114379882812\n",
            "tensor(107.6411, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 434, loss 107.64114379882812\n",
            "tensor(107.6411, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 435, loss 107.64114379882812\n",
            "tensor(107.6411, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 436, loss 107.64114379882812\n",
            "tensor(107.6411, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 437, loss 107.64114379882812\n",
            "tensor(107.6411, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 438, loss 107.64114379882812\n",
            "tensor(107.6411, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 439, loss 107.64114379882812\n",
            "tensor(107.6411, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 440, loss 107.64114379882812\n",
            "tensor(107.6411, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 441, loss 107.64114379882812\n",
            "tensor(107.6411, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 442, loss 107.64114379882812\n",
            "tensor(107.6411, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 443, loss 107.6411361694336\n",
            "tensor(107.6411, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 444, loss 107.6411361694336\n",
            "tensor(107.6411, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 445, loss 107.6411361694336\n",
            "tensor(107.6411, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 446, loss 107.6411361694336\n",
            "tensor(107.6411, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 447, loss 107.64112091064453\n",
            "tensor(107.6411, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 448, loss 107.64112091064453\n",
            "tensor(107.6411, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 449, loss 107.64112091064453\n",
            "tensor(107.6411, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 450, loss 107.64112091064453\n",
            "tensor(107.6411, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 451, loss 107.64112091064453\n",
            "tensor(107.6411, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 452, loss 107.64112091064453\n",
            "tensor(107.6411, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 453, loss 107.64112091064453\n",
            "tensor(107.6411, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 454, loss 107.64112091064453\n",
            "tensor(107.6411, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 455, loss 107.64112091064453\n",
            "tensor(107.6411, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 456, loss 107.64111328125\n",
            "tensor(107.6411, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 457, loss 107.64112091064453\n",
            "tensor(107.6411, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 458, loss 107.64111328125\n",
            "tensor(107.6411, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 459, loss 107.64111328125\n",
            "tensor(107.6411, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 460, loss 107.64109802246094\n",
            "tensor(107.6411, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 461, loss 107.64109802246094\n",
            "tensor(107.6411, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 462, loss 107.64111328125\n",
            "tensor(107.6411, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 463, loss 107.64109802246094\n",
            "tensor(107.6411, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 464, loss 107.64109802246094\n",
            "tensor(107.6411, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 465, loss 107.64109802246094\n",
            "tensor(107.6411, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 466, loss 107.64109802246094\n",
            "tensor(107.6411, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 467, loss 107.64109802246094\n",
            "tensor(107.6411, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 468, loss 107.64109802246094\n",
            "tensor(107.6411, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 469, loss 107.64109802246094\n",
            "tensor(107.6411, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 470, loss 107.64108276367188\n",
            "tensor(107.6411, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 471, loss 107.64109802246094\n",
            "tensor(107.6411, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 472, loss 107.64108276367188\n",
            "tensor(107.6411, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 473, loss 107.64108276367188\n",
            "tensor(107.6411, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 474, loss 107.64108276367188\n",
            "tensor(107.6411, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 475, loss 107.64107513427734\n",
            "tensor(107.6411, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 476, loss 107.64108276367188\n",
            "tensor(107.6411, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 477, loss 107.64107513427734\n",
            "tensor(107.6411, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 478, loss 107.64107513427734\n",
            "tensor(107.6411, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 479, loss 107.64107513427734\n",
            "tensor(107.6411, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 480, loss 107.64107513427734\n",
            "tensor(107.6411, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 481, loss 107.64107513427734\n",
            "tensor(107.6411, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 482, loss 107.64107513427734\n",
            "tensor(107.6411, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 483, loss 107.64107513427734\n",
            "tensor(107.6411, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 484, loss 107.64107513427734\n",
            "tensor(107.6411, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 485, loss 107.64105987548828\n",
            "tensor(107.6411, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 486, loss 107.64105987548828\n",
            "tensor(107.6411, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 487, loss 107.64105987548828\n",
            "tensor(107.6411, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 488, loss 107.64105987548828\n",
            "tensor(107.6411, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 489, loss 107.64105987548828\n",
            "tensor(107.6411, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 490, loss 107.64105987548828\n",
            "tensor(107.6410, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 491, loss 107.64104461669922\n",
            "tensor(107.6410, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 492, loss 107.64104461669922\n",
            "tensor(107.6410, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 493, loss 107.64104461669922\n",
            "tensor(107.6410, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 494, loss 107.64104461669922\n",
            "tensor(107.6410, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 495, loss 107.64104461669922\n",
            "tensor(107.6410, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 496, loss 107.64104461669922\n",
            "tensor(107.6410, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 497, loss 107.64104461669922\n",
            "tensor(107.6410, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 498, loss 107.64104461669922\n",
            "tensor(107.6410, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 499, loss 107.64104461669922\n",
            "tensor(107.6410, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 500, loss 107.64104461669922\n",
            "tensor(107.6410, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 501, loss 107.64103698730469\n",
            "tensor(107.6410, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 502, loss 107.64103698730469\n",
            "tensor(107.6410, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 503, loss 107.64103698730469\n",
            "tensor(107.6410, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 504, loss 107.64103698730469\n",
            "tensor(107.6410, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 505, loss 107.64103698730469\n",
            "tensor(107.6410, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 506, loss 107.64102172851562\n",
            "tensor(107.6410, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 507, loss 107.64102172851562\n",
            "tensor(107.6410, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 508, loss 107.64102172851562\n",
            "tensor(107.6410, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 509, loss 107.64102172851562\n",
            "tensor(107.6410, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 510, loss 107.64102172851562\n",
            "tensor(107.6410, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 511, loss 107.64102172851562\n",
            "tensor(107.6410, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 512, loss 107.64102172851562\n",
            "tensor(107.6410, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 513, loss 107.64102172851562\n",
            "tensor(107.6410, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 514, loss 107.64102172851562\n",
            "tensor(107.6410, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 515, loss 107.64102172851562\n",
            "tensor(107.6410, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 516, loss 107.64100646972656\n",
            "tensor(107.6410, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 517, loss 107.64100646972656\n",
            "tensor(107.6410, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 518, loss 107.64100646972656\n",
            "tensor(107.6410, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 519, loss 107.64100646972656\n",
            "tensor(107.6410, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 520, loss 107.64099884033203\n",
            "tensor(107.6410, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 521, loss 107.64099884033203\n",
            "tensor(107.6410, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 522, loss 107.64099884033203\n",
            "tensor(107.6410, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 523, loss 107.64099884033203\n",
            "tensor(107.6410, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 524, loss 107.64099884033203\n",
            "tensor(107.6410, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 525, loss 107.64099884033203\n",
            "tensor(107.6410, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 526, loss 107.64099884033203\n",
            "tensor(107.6410, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 527, loss 107.64099884033203\n",
            "tensor(107.6410, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 528, loss 107.64099884033203\n",
            "tensor(107.6410, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 529, loss 107.64099884033203\n",
            "tensor(107.6410, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 530, loss 107.64099884033203\n",
            "tensor(107.6410, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 531, loss 107.64098358154297\n",
            "tensor(107.6410, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 532, loss 107.64098358154297\n",
            "tensor(107.6410, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 533, loss 107.64098358154297\n",
            "tensor(107.6410, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 534, loss 107.64097595214844\n",
            "tensor(107.6410, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 535, loss 107.64098358154297\n",
            "tensor(107.6410, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 536, loss 107.64097595214844\n",
            "tensor(107.6410, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 537, loss 107.64097595214844\n",
            "tensor(107.6410, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 538, loss 107.64097595214844\n",
            "tensor(107.6410, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 539, loss 107.64097595214844\n",
            "tensor(107.6410, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 540, loss 107.64097595214844\n",
            "tensor(107.6410, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 541, loss 107.64097595214844\n",
            "tensor(107.6410, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 542, loss 107.64097595214844\n",
            "tensor(107.6410, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 543, loss 107.64097595214844\n",
            "tensor(107.6410, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 544, loss 107.64097595214844\n",
            "tensor(107.6410, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 545, loss 107.64097595214844\n",
            "tensor(107.6410, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 546, loss 107.64096069335938\n",
            "tensor(107.6410, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 547, loss 107.64096069335938\n",
            "tensor(107.6410, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 548, loss 107.64096069335938\n",
            "tensor(107.6410, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 549, loss 107.64096069335938\n",
            "tensor(107.6409, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 550, loss 107.64094543457031\n",
            "tensor(107.6409, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 551, loss 107.64094543457031\n",
            "tensor(107.6409, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 552, loss 107.64094543457031\n",
            "tensor(107.6409, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 553, loss 107.64094543457031\n",
            "tensor(107.6409, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 554, loss 107.64094543457031\n",
            "tensor(107.6409, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 555, loss 107.64094543457031\n",
            "tensor(107.6409, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 556, loss 107.64094543457031\n",
            "tensor(107.6409, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 557, loss 107.64094543457031\n",
            "tensor(107.6409, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 558, loss 107.64094543457031\n",
            "tensor(107.6409, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 559, loss 107.64094543457031\n",
            "tensor(107.6409, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 560, loss 107.64094543457031\n",
            "tensor(107.6409, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 561, loss 107.64094543457031\n",
            "tensor(107.6409, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 562, loss 107.64092254638672\n",
            "tensor(107.6409, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 563, loss 107.64092254638672\n",
            "tensor(107.6409, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 564, loss 107.64092254638672\n",
            "tensor(107.6409, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 565, loss 107.64092254638672\n",
            "tensor(107.6409, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 566, loss 107.64092254638672\n",
            "tensor(107.6409, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 567, loss 107.64092254638672\n",
            "tensor(107.6409, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 568, loss 107.64092254638672\n",
            "tensor(107.6409, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 569, loss 107.64092254638672\n",
            "tensor(107.6409, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 570, loss 107.64092254638672\n",
            "tensor(107.6409, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 571, loss 107.64092254638672\n",
            "tensor(107.6409, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 572, loss 107.64092254638672\n",
            "tensor(107.6409, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 573, loss 107.64092254638672\n",
            "tensor(107.6409, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 574, loss 107.64092254638672\n",
            "tensor(107.6409, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 575, loss 107.64090728759766\n",
            "tensor(107.6409, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 576, loss 107.64092254638672\n",
            "tensor(107.6409, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 577, loss 107.64089965820312\n",
            "tensor(107.6409, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 578, loss 107.64089965820312\n",
            "tensor(107.6409, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 579, loss 107.64089965820312\n",
            "tensor(107.6409, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 580, loss 107.64089965820312\n",
            "tensor(107.6409, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 581, loss 107.64089965820312\n",
            "tensor(107.6409, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 582, loss 107.64089965820312\n",
            "tensor(107.6409, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 583, loss 107.64089965820312\n",
            "tensor(107.6409, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 584, loss 107.64089965820312\n",
            "tensor(107.6409, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 585, loss 107.64089965820312\n",
            "tensor(107.6409, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 586, loss 107.64089965820312\n",
            "tensor(107.6409, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 587, loss 107.64088439941406\n",
            "tensor(107.6409, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 588, loss 107.64089965820312\n",
            "tensor(107.6409, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 589, loss 107.64088439941406\n",
            "tensor(107.6409, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 590, loss 107.64088439941406\n",
            "tensor(107.6409, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 591, loss 107.640869140625\n",
            "tensor(107.6409, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 592, loss 107.64088439941406\n",
            "tensor(107.6409, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 593, loss 107.640869140625\n",
            "tensor(107.6409, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 594, loss 107.640869140625\n",
            "tensor(107.6409, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 595, loss 107.640869140625\n",
            "tensor(107.6409, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 596, loss 107.640869140625\n",
            "tensor(107.6409, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 597, loss 107.640869140625\n",
            "tensor(107.6409, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 598, loss 107.640869140625\n",
            "tensor(107.6409, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 599, loss 107.640869140625\n",
            "tensor(107.6409, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 600, loss 107.640869140625\n",
            "tensor(107.6409, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 601, loss 107.640869140625\n",
            "tensor(107.6409, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 602, loss 107.640869140625\n",
            "tensor(107.6409, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 603, loss 107.640869140625\n",
            "tensor(107.6409, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 604, loss 107.640869140625\n",
            "tensor(107.6409, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 605, loss 107.640869140625\n",
            "tensor(107.6408, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 606, loss 107.6408462524414\n",
            "tensor(107.6408, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 607, loss 107.6408462524414\n",
            "tensor(107.6408, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 608, loss 107.6408462524414\n",
            "tensor(107.6408, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 609, loss 107.6408462524414\n",
            "tensor(107.6408, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 610, loss 107.6408462524414\n",
            "tensor(107.6408, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 611, loss 107.6408462524414\n",
            "tensor(107.6408, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 612, loss 107.6408462524414\n",
            "tensor(107.6408, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 613, loss 107.6408462524414\n",
            "tensor(107.6408, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 614, loss 107.6408462524414\n",
            "tensor(107.6408, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 615, loss 107.6408462524414\n",
            "tensor(107.6408, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 616, loss 107.6408462524414\n",
            "tensor(107.6408, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 617, loss 107.6408462524414\n",
            "tensor(107.6408, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 618, loss 107.64083099365234\n",
            "tensor(107.6408, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 619, loss 107.6408462524414\n",
            "tensor(107.6408, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 620, loss 107.64082336425781\n",
            "tensor(107.6408, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 621, loss 107.64083099365234\n",
            "tensor(107.6408, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 622, loss 107.64082336425781\n",
            "tensor(107.6408, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 623, loss 107.64082336425781\n",
            "tensor(107.6408, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 624, loss 107.64082336425781\n",
            "tensor(107.6408, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 625, loss 107.64082336425781\n",
            "tensor(107.6408, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 626, loss 107.64082336425781\n",
            "tensor(107.6408, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 627, loss 107.64082336425781\n",
            "tensor(107.6408, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 628, loss 107.64082336425781\n",
            "tensor(107.6408, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 629, loss 107.64082336425781\n",
            "tensor(107.6408, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 630, loss 107.64082336425781\n",
            "tensor(107.6408, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 631, loss 107.64082336425781\n",
            "tensor(107.6408, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 632, loss 107.64082336425781\n",
            "tensor(107.6408, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 633, loss 107.64080810546875\n",
            "tensor(107.6408, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 634, loss 107.64082336425781\n",
            "tensor(107.6408, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 635, loss 107.64080047607422\n",
            "tensor(107.6408, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 636, loss 107.64080047607422\n",
            "tensor(107.6408, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 637, loss 107.64080047607422\n",
            "tensor(107.6408, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 638, loss 107.64080047607422\n",
            "tensor(107.6408, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 639, loss 107.64080047607422\n",
            "tensor(107.6408, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 640, loss 107.64080047607422\n",
            "tensor(107.6408, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 641, loss 107.64080047607422\n",
            "tensor(107.6408, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 642, loss 107.64080047607422\n",
            "tensor(107.6408, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 643, loss 107.64080047607422\n",
            "tensor(107.6408, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 644, loss 107.64080047607422\n",
            "tensor(107.6408, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 645, loss 107.64080047607422\n",
            "tensor(107.6408, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 646, loss 107.64080047607422\n",
            "tensor(107.6408, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 647, loss 107.64080047607422\n",
            "tensor(107.6408, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 648, loss 107.6407699584961\n",
            "tensor(107.6408, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 649, loss 107.6407699584961\n",
            "tensor(107.6408, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 650, loss 107.64080047607422\n",
            "tensor(107.6408, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 651, loss 107.6407699584961\n",
            "tensor(107.6408, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 652, loss 107.6407699584961\n",
            "tensor(107.6408, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 653, loss 107.6407699584961\n",
            "tensor(107.6408, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 654, loss 107.6407699584961\n",
            "tensor(107.6408, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 655, loss 107.6407699584961\n",
            "tensor(107.6408, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 656, loss 107.6407699584961\n",
            "tensor(107.6408, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 657, loss 107.6407699584961\n",
            "tensor(107.6408, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 658, loss 107.6407699584961\n",
            "tensor(107.6408, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 659, loss 107.6407699584961\n",
            "tensor(107.6408, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 660, loss 107.6407699584961\n",
            "tensor(107.6408, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 661, loss 107.6407699584961\n",
            "tensor(107.6408, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 662, loss 107.6407699584961\n",
            "tensor(107.6408, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 663, loss 107.6407699584961\n",
            "tensor(107.6408, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 664, loss 107.64076232910156\n",
            "tensor(107.6407, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 665, loss 107.6407470703125\n",
            "tensor(107.6408, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 666, loss 107.64076232910156\n",
            "tensor(107.6407, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 667, loss 107.6407470703125\n",
            "tensor(107.6407, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 668, loss 107.6407470703125\n",
            "tensor(107.6407, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 669, loss 107.6407470703125\n",
            "tensor(107.6407, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 670, loss 107.6407470703125\n",
            "tensor(107.6407, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 671, loss 107.6407470703125\n",
            "tensor(107.6407, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 672, loss 107.6407470703125\n",
            "tensor(107.6407, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 673, loss 107.6407470703125\n",
            "tensor(107.6407, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 674, loss 107.6407470703125\n",
            "tensor(107.6407, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 675, loss 107.6407470703125\n",
            "tensor(107.6407, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 676, loss 107.6407470703125\n",
            "tensor(107.6407, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 677, loss 107.64073181152344\n",
            "tensor(107.6407, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 678, loss 107.6407241821289\n",
            "tensor(107.6407, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 679, loss 107.6407241821289\n",
            "tensor(107.6407, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 680, loss 107.6407241821289\n",
            "tensor(107.6407, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 681, loss 107.6407241821289\n",
            "tensor(107.6407, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 682, loss 107.6407241821289\n",
            "tensor(107.6407, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 683, loss 107.6407241821289\n",
            "tensor(107.6407, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 684, loss 107.6407241821289\n",
            "tensor(107.6407, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 685, loss 107.6407241821289\n",
            "tensor(107.6407, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 686, loss 107.6407241821289\n",
            "tensor(107.6407, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 687, loss 107.6407241821289\n",
            "tensor(107.6407, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 688, loss 107.6407241821289\n",
            "tensor(107.6407, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 689, loss 107.6407241821289\n",
            "tensor(107.6407, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 690, loss 107.6407241821289\n",
            "tensor(107.6407, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 691, loss 107.6407241821289\n",
            "tensor(107.6407, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 692, loss 107.6407241821289\n",
            "tensor(107.6407, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 693, loss 107.6407241821289\n",
            "tensor(107.6407, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 694, loss 107.64070892333984\n",
            "tensor(107.6407, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 695, loss 107.64069366455078\n",
            "tensor(107.6407, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 696, loss 107.64069366455078\n",
            "tensor(107.6407, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 697, loss 107.64069366455078\n",
            "tensor(107.6407, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 698, loss 107.64069366455078\n",
            "tensor(107.6407, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 699, loss 107.64069366455078\n",
            "tensor(107.6407, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 700, loss 107.64069366455078\n",
            "tensor(107.6407, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 701, loss 107.64069366455078\n",
            "tensor(107.6407, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 702, loss 107.64069366455078\n",
            "tensor(107.6407, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 703, loss 107.64069366455078\n",
            "tensor(107.6407, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 704, loss 107.64069366455078\n",
            "tensor(107.6407, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 705, loss 107.64069366455078\n",
            "tensor(107.6407, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 706, loss 107.64069366455078\n",
            "tensor(107.6407, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 707, loss 107.64069366455078\n",
            "tensor(107.6407, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 708, loss 107.64068603515625\n",
            "tensor(107.6407, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 709, loss 107.64067077636719\n",
            "tensor(107.6407, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 710, loss 107.64068603515625\n",
            "tensor(107.6407, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 711, loss 107.64067077636719\n",
            "tensor(107.6407, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 712, loss 107.64067077636719\n",
            "tensor(107.6407, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 713, loss 107.64067077636719\n",
            "tensor(107.6407, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 714, loss 107.64067077636719\n",
            "tensor(107.6407, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 715, loss 107.64067077636719\n",
            "tensor(107.6407, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 716, loss 107.64067077636719\n",
            "tensor(107.6407, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 717, loss 107.64067077636719\n",
            "tensor(107.6407, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 718, loss 107.64067077636719\n",
            "tensor(107.6407, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 719, loss 107.64067077636719\n",
            "tensor(107.6407, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 720, loss 107.64067077636719\n",
            "tensor(107.6407, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 721, loss 107.64067077636719\n",
            "tensor(107.6407, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 722, loss 107.64066314697266\n",
            "tensor(107.6407, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 723, loss 107.64067077636719\n",
            "tensor(107.6406, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 724, loss 107.6406478881836\n",
            "tensor(107.6406, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 725, loss 107.6406478881836\n",
            "tensor(107.6406, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 726, loss 107.6406478881836\n",
            "tensor(107.6406, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 727, loss 107.6406478881836\n",
            "tensor(107.6406, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 728, loss 107.6406478881836\n",
            "tensor(107.6406, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 729, loss 107.6406478881836\n",
            "tensor(107.6406, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 730, loss 107.6406478881836\n",
            "tensor(107.6406, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 731, loss 107.6406478881836\n",
            "tensor(107.6406, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 732, loss 107.6406478881836\n",
            "tensor(107.6406, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 733, loss 107.6406478881836\n",
            "tensor(107.6406, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 734, loss 107.6406478881836\n",
            "tensor(107.6406, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 735, loss 107.6406478881836\n",
            "tensor(107.6406, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 736, loss 107.6406478881836\n",
            "tensor(107.6406, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 737, loss 107.64063262939453\n",
            "tensor(107.6406, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 738, loss 107.6406478881836\n",
            "tensor(107.6406, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 739, loss 107.640625\n",
            "tensor(107.6406, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 740, loss 107.640625\n",
            "tensor(107.6406, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 741, loss 107.640625\n",
            "tensor(107.6406, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 742, loss 107.640625\n",
            "tensor(107.6406, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 743, loss 107.640625\n",
            "tensor(107.6406, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 744, loss 107.640625\n",
            "tensor(107.6406, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 745, loss 107.640625\n",
            "tensor(107.6406, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 746, loss 107.640625\n",
            "tensor(107.6406, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 747, loss 107.640625\n",
            "tensor(107.6406, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 748, loss 107.640625\n",
            "tensor(107.6406, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 749, loss 107.640625\n",
            "tensor(107.6406, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 750, loss 107.64060974121094\n",
            "tensor(107.6406, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 751, loss 107.640625\n",
            "tensor(107.6406, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 752, loss 107.64060974121094\n",
            "tensor(107.6406, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 753, loss 107.64060974121094\n",
            "tensor(107.6406, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 754, loss 107.64060974121094\n",
            "tensor(107.6406, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 755, loss 107.64060974121094\n",
            "tensor(107.6406, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 756, loss 107.64059448242188\n",
            "tensor(107.6406, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 757, loss 107.64059448242188\n",
            "tensor(107.6406, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 758, loss 107.64059448242188\n",
            "tensor(107.6406, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 759, loss 107.64059448242188\n",
            "tensor(107.6406, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 760, loss 107.64059448242188\n",
            "tensor(107.6406, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 761, loss 107.64059448242188\n",
            "tensor(107.6406, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 762, loss 107.64059448242188\n",
            "tensor(107.6406, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 763, loss 107.64059448242188\n",
            "tensor(107.6406, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 764, loss 107.64058685302734\n",
            "tensor(107.6406, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 765, loss 107.64058685302734\n",
            "tensor(107.6406, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 766, loss 107.64058685302734\n",
            "tensor(107.6406, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 767, loss 107.64057159423828\n",
            "tensor(107.6406, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 768, loss 107.64057159423828\n",
            "tensor(107.6406, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 769, loss 107.64057159423828\n",
            "tensor(107.6406, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 770, loss 107.64057159423828\n",
            "tensor(107.6406, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 771, loss 107.64057159423828\n",
            "tensor(107.6406, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 772, loss 107.64057159423828\n",
            "tensor(107.6406, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 773, loss 107.64057159423828\n",
            "tensor(107.6406, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 774, loss 107.64057159423828\n",
            "tensor(107.6406, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 775, loss 107.64057159423828\n",
            "tensor(107.6406, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 776, loss 107.64057159423828\n",
            "tensor(107.6406, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 777, loss 107.64057159423828\n",
            "tensor(107.6406, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 778, loss 107.64057159423828\n",
            "tensor(107.6406, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 779, loss 107.64057159423828\n",
            "tensor(107.6406, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 780, loss 107.64055633544922\n",
            "tensor(107.6406, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 781, loss 107.64057159423828\n",
            "tensor(107.6405, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 782, loss 107.64054870605469\n",
            "tensor(107.6405, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 783, loss 107.64054870605469\n",
            "tensor(107.6405, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 784, loss 107.64054870605469\n",
            "tensor(107.6405, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 785, loss 107.64054870605469\n",
            "tensor(107.6405, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 786, loss 107.64054870605469\n",
            "tensor(107.6405, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 787, loss 107.64054870605469\n",
            "tensor(107.6405, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 788, loss 107.64054870605469\n",
            "tensor(107.6405, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 789, loss 107.64054870605469\n",
            "tensor(107.6405, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 790, loss 107.64054870605469\n",
            "tensor(107.6405, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 791, loss 107.64054870605469\n",
            "tensor(107.6405, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 792, loss 107.64054870605469\n",
            "tensor(107.6405, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 793, loss 107.64054870605469\n",
            "tensor(107.6405, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 794, loss 107.64054870605469\n",
            "tensor(107.6405, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 795, loss 107.64053344726562\n",
            "tensor(107.6405, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 796, loss 107.64054870605469\n",
            "tensor(107.6405, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 797, loss 107.64054870605469\n",
            "tensor(107.6405, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 798, loss 107.6405258178711\n",
            "tensor(107.6405, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 799, loss 107.6405258178711\n",
            "tensor(107.6405, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 800, loss 107.6405258178711\n",
            "tensor(107.6405, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 801, loss 107.6405258178711\n",
            "tensor(107.6405, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 802, loss 107.6405258178711\n",
            "tensor(107.6405, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 803, loss 107.6405258178711\n",
            "tensor(107.6405, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 804, loss 107.6405258178711\n",
            "tensor(107.6405, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 805, loss 107.6405258178711\n",
            "tensor(107.6405, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 806, loss 107.6405258178711\n",
            "tensor(107.6405, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 807, loss 107.6405258178711\n",
            "tensor(107.6405, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 808, loss 107.6405258178711\n",
            "tensor(107.6405, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 809, loss 107.64051055908203\n",
            "tensor(107.6405, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 810, loss 107.64051055908203\n",
            "tensor(107.6405, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 811, loss 107.64049530029297\n",
            "tensor(107.6405, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 812, loss 107.6405258178711\n",
            "tensor(107.6405, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 813, loss 107.64051055908203\n",
            "tensor(107.6405, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 814, loss 107.64049530029297\n",
            "tensor(107.6405, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 815, loss 107.64049530029297\n",
            "tensor(107.6405, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 816, loss 107.64049530029297\n",
            "tensor(107.6405, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 817, loss 107.64049530029297\n",
            "tensor(107.6405, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 818, loss 107.64049530029297\n",
            "tensor(107.6405, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 819, loss 107.64049530029297\n",
            "tensor(107.6405, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 820, loss 107.64049530029297\n",
            "tensor(107.6405, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 821, loss 107.64049530029297\n",
            "tensor(107.6405, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 822, loss 107.64049530029297\n",
            "tensor(107.6405, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 823, loss 107.64049530029297\n",
            "tensor(107.6405, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 824, loss 107.64049530029297\n",
            "tensor(107.6405, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 825, loss 107.64048767089844\n",
            "tensor(107.6405, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 826, loss 107.64048767089844\n",
            "tensor(107.6405, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 827, loss 107.64047241210938\n",
            "tensor(107.6405, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 828, loss 107.64047241210938\n",
            "tensor(107.6405, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 829, loss 107.64047241210938\n",
            "tensor(107.6405, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 830, loss 107.64047241210938\n",
            "tensor(107.6405, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 831, loss 107.64047241210938\n",
            "tensor(107.6405, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 832, loss 107.64047241210938\n",
            "tensor(107.6405, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 833, loss 107.64047241210938\n",
            "tensor(107.6405, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 834, loss 107.64047241210938\n",
            "tensor(107.6405, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 835, loss 107.64047241210938\n",
            "tensor(107.6405, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 836, loss 107.64047241210938\n",
            "tensor(107.6405, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 837, loss 107.64047241210938\n",
            "tensor(107.6405, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 838, loss 107.64047241210938\n",
            "tensor(107.6405, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 839, loss 107.64045715332031\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 840, loss 107.64044952392578\n",
            "tensor(107.6405, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 841, loss 107.64045715332031\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 842, loss 107.64044952392578\n",
            "tensor(107.6405, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 843, loss 107.64045715332031\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 844, loss 107.64044952392578\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 845, loss 107.64044952392578\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 846, loss 107.64044952392578\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 847, loss 107.64044952392578\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 848, loss 107.64044952392578\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 849, loss 107.64044952392578\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 850, loss 107.64044952392578\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 851, loss 107.64044952392578\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 852, loss 107.64044952392578\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 853, loss 107.64043426513672\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 854, loss 107.64044952392578\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 855, loss 107.64043426513672\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 856, loss 107.64041900634766\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 857, loss 107.64041900634766\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 858, loss 107.64043426513672\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 859, loss 107.64041900634766\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 860, loss 107.64041900634766\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 861, loss 107.64041900634766\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 862, loss 107.64041900634766\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 863, loss 107.64041900634766\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 864, loss 107.64041900634766\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 865, loss 107.64041900634766\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 866, loss 107.64041900634766\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 867, loss 107.64041900634766\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 868, loss 107.64041137695312\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 869, loss 107.64041137695312\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 870, loss 107.64041137695312\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 871, loss 107.64039611816406\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 872, loss 107.64041137695312\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 873, loss 107.64039611816406\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 874, loss 107.64039611816406\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 875, loss 107.64039611816406\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 876, loss 107.64039611816406\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 877, loss 107.64039611816406\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 878, loss 107.64039611816406\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 879, loss 107.64039611816406\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 880, loss 107.64039611816406\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 881, loss 107.64038848876953\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 882, loss 107.64038848876953\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 883, loss 107.64038848876953\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 884, loss 107.64038848876953\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 885, loss 107.64038848876953\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 886, loss 107.64038848876953\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 887, loss 107.64037322998047\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 888, loss 107.64037322998047\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 889, loss 107.64037322998047\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 890, loss 107.64037322998047\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 891, loss 107.64037322998047\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 892, loss 107.64037322998047\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 893, loss 107.64037322998047\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 894, loss 107.64037322998047\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 895, loss 107.64037322998047\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 896, loss 107.64037322998047\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 897, loss 107.64037322998047\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 898, loss 107.6403579711914\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 899, loss 107.64035034179688\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 900, loss 107.64035034179688\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 901, loss 107.64035034179688\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 902, loss 107.64035034179688\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 903, loss 107.64035034179688\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 904, loss 107.64035034179688\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 905, loss 107.64035034179688\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 906, loss 107.64035034179688\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 907, loss 107.64035034179688\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 908, loss 107.64035034179688\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 909, loss 107.64035034179688\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 910, loss 107.64035034179688\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 911, loss 107.64035034179688\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 912, loss 107.64035034179688\n",
            "tensor(107.6404, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 913, loss 107.64035034179688\n",
            "tensor(107.6403, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 914, loss 107.64031982421875\n",
            "tensor(107.6403, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 915, loss 107.64033508300781\n",
            "tensor(107.6403, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 916, loss 107.64031982421875\n",
            "tensor(107.6403, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 917, loss 107.64031982421875\n",
            "tensor(107.6403, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 918, loss 107.64031982421875\n",
            "tensor(107.6403, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 919, loss 107.64031982421875\n",
            "tensor(107.6403, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 920, loss 107.64031982421875\n",
            "tensor(107.6403, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 921, loss 107.64031982421875\n",
            "tensor(107.6403, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 922, loss 107.64031982421875\n",
            "tensor(107.6403, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 923, loss 107.64031982421875\n",
            "tensor(107.6403, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 924, loss 107.64031982421875\n",
            "tensor(107.6403, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 925, loss 107.64031982421875\n",
            "tensor(107.6403, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 926, loss 107.64031982421875\n",
            "tensor(107.6403, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 927, loss 107.64031219482422\n",
            "tensor(107.6403, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 928, loss 107.64029693603516\n",
            "tensor(107.6403, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 929, loss 107.64029693603516\n",
            "tensor(107.6403, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 930, loss 107.64029693603516\n",
            "tensor(107.6403, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 931, loss 107.64029693603516\n",
            "tensor(107.6403, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 932, loss 107.64029693603516\n",
            "tensor(107.6403, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 933, loss 107.64029693603516\n",
            "tensor(107.6403, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 934, loss 107.64029693603516\n",
            "tensor(107.6403, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 935, loss 107.64029693603516\n",
            "tensor(107.6403, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 936, loss 107.64029693603516\n",
            "tensor(107.6403, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 937, loss 107.64029693603516\n",
            "tensor(107.6403, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 938, loss 107.64029693603516\n",
            "tensor(107.6403, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 939, loss 107.64029693603516\n",
            "tensor(107.6403, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 940, loss 107.64029693603516\n",
            "tensor(107.6403, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 941, loss 107.6402816772461\n",
            "tensor(107.6403, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 942, loss 107.64027404785156\n",
            "tensor(107.6403, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 943, loss 107.6402816772461\n",
            "tensor(107.6403, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 944, loss 107.64027404785156\n",
            "tensor(107.6403, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 945, loss 107.64027404785156\n",
            "tensor(107.6403, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 946, loss 107.64027404785156\n",
            "tensor(107.6403, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 947, loss 107.64027404785156\n",
            "tensor(107.6403, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 948, loss 107.64027404785156\n",
            "tensor(107.6403, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 949, loss 107.64027404785156\n",
            "tensor(107.6403, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 950, loss 107.64027404785156\n",
            "tensor(107.6403, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 951, loss 107.64027404785156\n",
            "tensor(107.6403, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 952, loss 107.64027404785156\n",
            "tensor(107.6403, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 953, loss 107.6402587890625\n",
            "tensor(107.6403, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 954, loss 107.6402587890625\n",
            "tensor(107.6403, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 955, loss 107.6402587890625\n",
            "tensor(107.6402, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 956, loss 107.64024353027344\n",
            "tensor(107.6402, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 957, loss 107.64024353027344\n",
            "tensor(107.6402, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 958, loss 107.64024353027344\n",
            "tensor(107.6402, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 959, loss 107.64024353027344\n",
            "tensor(107.6402, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 960, loss 107.64024353027344\n",
            "tensor(107.6402, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 961, loss 107.64024353027344\n",
            "tensor(107.6402, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 962, loss 107.64024353027344\n",
            "tensor(107.6402, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 963, loss 107.64024353027344\n",
            "tensor(107.6402, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 964, loss 107.64024353027344\n",
            "tensor(107.6402, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 965, loss 107.64024353027344\n",
            "tensor(107.6402, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 966, loss 107.64024353027344\n",
            "tensor(107.6402, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 967, loss 107.64024353027344\n",
            "tensor(107.6402, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 968, loss 107.64024353027344\n",
            "tensor(107.6402, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 969, loss 107.64024353027344\n",
            "tensor(107.6402, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 970, loss 107.6402359008789\n",
            "tensor(107.6402, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 971, loss 107.64022064208984\n",
            "tensor(107.6402, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 972, loss 107.6402359008789\n",
            "tensor(107.6402, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 973, loss 107.64022064208984\n",
            "tensor(107.6402, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 974, loss 107.64022064208984\n",
            "tensor(107.6402, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 975, loss 107.64022064208984\n",
            "tensor(107.6402, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 976, loss 107.64022064208984\n",
            "tensor(107.6402, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 977, loss 107.64022064208984\n",
            "tensor(107.6402, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 978, loss 107.64022064208984\n",
            "tensor(107.6402, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 979, loss 107.64022064208984\n",
            "tensor(107.6402, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 980, loss 107.64022064208984\n",
            "tensor(107.6402, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 981, loss 107.64022064208984\n",
            "tensor(107.6402, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 982, loss 107.64022064208984\n",
            "tensor(107.6402, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 983, loss 107.64022064208984\n",
            "tensor(107.6402, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 984, loss 107.64022064208984\n",
            "tensor(107.6402, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 985, loss 107.64019775390625\n",
            "tensor(107.6402, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 986, loss 107.64021301269531\n",
            "tensor(107.6402, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 987, loss 107.64019775390625\n",
            "tensor(107.6402, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 988, loss 107.64019775390625\n",
            "tensor(107.6402, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 989, loss 107.64019775390625\n",
            "tensor(107.6402, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 990, loss 107.64019775390625\n",
            "tensor(107.6402, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 991, loss 107.64019775390625\n",
            "tensor(107.6402, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 992, loss 107.64019775390625\n",
            "tensor(107.6402, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 993, loss 107.64019775390625\n",
            "tensor(107.6402, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 994, loss 107.64019775390625\n",
            "tensor(107.6402, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 995, loss 107.64019775390625\n",
            "tensor(107.6402, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 996, loss 107.64019775390625\n",
            "tensor(107.6402, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 997, loss 107.64019775390625\n",
            "tensor(107.6402, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 998, loss 107.64019775390625\n",
            "tensor(107.6402, device='cuda:0', grad_fn=<MseLossBackward0>)\n",
            "epoch 999, loss 107.64019775390625\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ZPmtQ3cKTkNk"
      },
      "source": [
        "#### Plot Predicted Height vs Actual Height"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "EVyQVHzLTlZI",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 389
        },
        "outputId": "41ca00fe-7a8b-42a9-8a62-5ef5182168bf"
      },
      "source": [
        "with torch.no_grad():\n",
        "    predicted = model(Variable(torch.from_numpy(x_test).cuda())).cpu().data.numpy()\n",
        "    print(predicted)\n",
        "\n",
        "plt.clf()\n",
        "plt.plot(x_test, y_test, 'go', label='Actual Height', alpha=0.5)\n",
        "plt.plot(x_test, predicted, '--', label='Predicted Height', alpha=0.5)\n",
        "plt.legend(loc='best')\n",
        "plt.show()"
      ],
      "execution_count": 23,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[[134.5343 ]\n",
            " [128.31352]\n",
            " [125.81574]\n",
            " ...\n",
            " [126.89599]\n",
            " [129.54024]\n",
            " [126.70605]]\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "APE4A-xbW906"
      },
      "source": [
        "#### Print RMSE"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "WPRv8yfHW_tm",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "d5cc6ccf-2dab-4a35-de7e-de0331f8e04f"
      },
      "source": [
        "#RMSE Root Mean Squared Error\n",
        "rms = sqrt(mean_squared_error(x_test, predicted))\n",
        "rms"
      ],
      "execution_count": 24,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "59.26140037341971"
            ]
          },
          "metadata": {},
          "execution_count": 24
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Example Coefficients"
      ],
      "metadata": {
        "id": "VvhscvJf2chr"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import numpy as np\n",
        "from sklearn.linear_model import LinearRegression\n",
        "X = np.array([[1, 1], [1, 2], [2, 2], [2, 3]])\n",
        "# y = 1 * x_0 + 2 * x_1 + 3\n",
        "y = np.dot(X, np.array([1, 2])) + 3\n",
        "reg = LinearRegression().fit(X, y)"
      ],
      "metadata": {
        "id": "H_JIxdr42bwh"
      },
      "execution_count": 35,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        ""
      ],
      "metadata": {
        "id": "LG-85B6T27CL"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "reg.score(X, y)\n",
        "reg.coef_\n",
        "reg.intercept_\n",
        "reg.predict(np.array([[3, 5]]))"
      ],
      "metadata": {
        "id": "ub05HzZa2qtM",
        "outputId": "42aa8646-fa57-4059-8534-83003d3af9b9",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "execution_count": 37,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([16.])"
            ]
          },
          "metadata": {},
          "execution_count": 37
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        ""
      ],
      "metadata": {
        "id": "duqncT3y2sbM"
      },
      "execution_count": null,
      "outputs": []
    }
  ]
}